Post1
This is a special thread, please read Post 2 below, which contains instructions for use.
 Read The Great Library: a hierarchical structure for further information about this type of threads.

Diplomats and Spies

4.2	Stacked units cannot be bribed or sabotaged. A unit is stacked if it share one square with ANY other(s) (overland, naval, aerial) unit.
4.3	Dip/spy may be bribed the same way as other units.
4.4	You get a non-unit if you bribe a unit that is closer to an enemy city than any your city.
4.5	Dip/spy moves to the point where bribed unit stays and loses only corresponding number of movement points. Bribed unit is avilable immediately and has full number of movement points.
4.6	Unit residing in a city cannot be bribed (i. e. cannot be bribed separately, only with all city).
	Formula for bribing units:
4.1	GR = DF * GF * S
	Where   GR = gold required to bribe    S = shield cost of the unit (see 4.13)   DF = 37.5 / (D+2)     (distance factor)   GF = 1+G / 750        (gold factor)   G = gold of civ that is bribed   D = the distance between unit that is bribed and his capital (see 4.10; see 4.31-4.35);
	Notes:
4.10	Maximal values for the distance "D":
	max. D  government of the civ of dip/spy 10 16  communism  other
	(Government (communism) of a civ whose units are bribed has no effect for bribing of units. (see 3.11) )
4.11	Unlike bribing cities, the cost for bribing unit is the same for dip/spy and doesn't depend on veteran status.
4.12	If capital does not exist then distance is always considered as maximal (16 or 10).
4.13	The cost to bribe units with a role=5 (settler, engineer) is double that of other units.
4.20	Bribing barbarian units: Formula 4.1 applies: there is no barbarian gold (G=0) and no barbarian capital.
	Way to measure distance between a unit and the capital (use in both bribing-formulas (3.1 and 4.1)):
4.31	Imagine Civ2-map turned by 45 degrees. Then you see that map is a simple grid of squares. Then use 4.32 or 4.33 (both are equal).
4.32	Take the "rectangle" defined by the position of the unit and the position of the city (all sides of rectangle are built by squares connected by lines, not by by edges). distance = (length of longer side of rectangle) + 0.5 * (length of shorter side of rectangle) Only integer part of distance is considered.
4.33	Take the "rhomboid" defined by the position of the unit and the position of the city (two sides of rhomboid are built by squares connected by lines ("straight sides" ), other two sides are built by squares connected by edges ("diagonal sides" ); note: the rhomboid is included in the rectangle that came up in 4.33). distance = (length of straight side of rhomboid) + 1.5 * (length of diagonal side of rhomboid) . Only integer part of distance is considered.
4.35	This method of measuring the distance between two squares is valid for all Civilization.
4.4	Help for people weak in mathematics:
	Distance 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 unit 125 93.5 75 62.5 53.5 46.5 41.5 37.5 34 31 28.5 26.5 25 23 22 20.5 city (dip) N/A 200 166 142 125 11 100 90 83 76 71 66 62 58 55 52 city (spy) N/A 167 139 119 105 93 84 75 70 64 60 55 52 50 46 44 city (vet spy) N/A 134 111 95 84 74 67 60 56 51 48 44 42 39 37 35
	2nd row applies to formula 4.1 (10 shield-unit, 0 gold), 3rd to 5th rows apply to 3.1 (1-sized city, 0 gold).

5.1

It is not true that

There are old forgotten (but very helpful) threads in Apolyton and it is very tedious if you try to extract a concrete information (I am experienced well with dips/spies).
I think a big part of Apolyton could be managed the way I suggest hereafter and then there wouldn't be so large redundancy of information. A few (10, 20) threads of similar fashion would span all questions (answered or unanswered) about Civ2 facts. To arrange strategies would be harder, I think they should be discussed ordinarily.

There is my idea how to manage this thread:

All wisdom will be concentraced to the main (first) post.

You will
post any suggestions :

to add new wisdom (will be marked green), to mark/unmark insecure or unsettled wisdom (will be marked red), to put new questions that should be answered (will be marked blue), remarks to mistakes, opacity and my weak english

and discuss them here.

I will
integrate your suggestions into the main post.

Note: There's no conflict between your summary and the Scrolls of Wisdom with regard to the veteran spy factor.

12.32 Attack "give primary target"
a) Basic example: attack of city walls in a capital city by vet spies.
Any attack is independent from the previous one (like "heads and tails"). Probability of success p1=0,5.
If the attack was successful, your vet spy might be captured afterwards. Probability of escaping p2=0,5.
b) All other examples derive from that one (but the probabilities differ). Testing with larger samples would be necessary to make sure of the exact values.
c) Meanwhile, I suppose that:
the probability of escaping after successful attack is close to
p2=0,5 for vet spies
p2=0,25 for non vet spies;
the probability of success is close to
p1=0,5 for vet spies attacking walls in a capital city;
the probability of success is higher when attacking improvements other than walls, and non capital cities;
the probability of success is much lower when attacking with non vet spies.
d)Hence it is advisable to attack with numerous vet spies using this "give primary target" method.
e)With non vet spies, it is generally advisable to choose the "use her judgment" method (especially if there are few improvements inside the city).

I suggest two changes to the formula 4.1:

I made an interesting discovery. Diplomats can subvert a city if there are no units in it.

La Fayette,
3.1 Did you read Post 21?? IMHO it clarifies all.

quote: Originally posted by La Fayette Let me know if there is anything else I could or should do.

3.13

quote: Originally posted by La Fayette (unfortunately I cannot send any save game since I play at home and post at my office with no link in between).

4.51 I suppose William Keenan and me solved the problem. See Post 24 and the appropriate part of Post 1.
Edit: sorry, I deleted the Post 24. See post 1.

Yes, post 20 is complete, I suppose it was my mistake, I wanted to point to the post 19. I didn't include it because I think there still remains some error in our tests: results are very strange.

quote: Originally posted by Slow Thinker on 03-20-2001 08:14 PM Winkler claims veteran status of both spy and diplomat affects bribing cost. I am prepared to remove a red note about Winkler at 3.1 ("5/6 or 2/3 if a spy or veteran spy is bribing") if anybody will confirm I am right and only spy is affected by vet status when bribing cities.

I have confirmed it. Winkler was wrong.

quote: I proved the previous owner thing. I agree. Is this expressed correctly (you know, my english...)? if dip/spy's civ was the last owner of the city before the curent owner. (If the civ founded it and it has changed hands several times, then no discount applies.)

I would just say: 1/2 if bribing civ was the previous owner.

quote: Originally posted by William Keenan I have only tested and verified the communism distance 10 factor with units, not cities. I will verify it for cities this weekend.

But I may not to be able to imagine your idea. Can you be more concrete? Or can you adjust a part of Post 1?

GR = S*(G+1000)/(D+3)

GR = Gold Required
G = gold of civ that is bribed
D = the distance between dip/spy and the capital of civ that is bribed

No need to have extra variables added in for no reason